We have continued studies of the mechanisms involved in activity dependent synapse elimination at the mouse neuromuscular junction in vitro. Thrombin and the thrombin receptor are essential for this elimination and the elimination is blocked by a specific protein kinase C inhibitor calphostin C, as well as the more general kinase inhibitors, staurosporine and H-7. The protein kinase A inhibitor, H-89, has no blocking effect on the activity dependent synapse elimination process. b. We studied expression of mRNA for prothrombin (PT), thrombin receptor (ThR), and the endogenous protease inhibitor protease-nexin 1 (PN-1) during development and after denervation of adult mouse muscle. PT and ThR were markedly down regulated during the first three weeks after birth, but PN-1 showed a transient decrease during the period of maximal naturally occurring synapse elimination (postnatal days 10-15). PT and ThR were substantially increased following denervation while PN-1 decreased to 50% of control. The decrease in PN-1 during the elimination period may result in an enhanced thrombin (or other proteolytic) activity, which our in vitro results indicate would induce synapse loss. c. Experiments with chick embryos indicate that both a trophic factor (brain derived neurotrophic factor or BDNF) and serine proteases (probably thrombin) are involved in the loss of the ipsilateral retinotectal projection in this preparation. Systemic hirudin or PN-1 produce a significant slowing of the loss of projection while BDNF injected in the eye prevents loss completely. d. Further experiments have been done describing retinal axon growth cone behavior as the growth cones encounter target cells in the chick tectum or mouse colliculus and analyzing the cellular and topographic distribution of cell surface guidance molecules. We have developed a model incorporating the cellular distribution of mRNA expression for ephrin-A2 and ephrin-A5 (known repulsive guidance molecules) with growth cone behaviors which accounts for differences in the patterns of retinotarget development in the mouse and chick.